The most common sources of electoral data on-line are a country's election agency, its national statistics institute, or the legislature. However, these links are often maintained for a small period of time. In addition, in many small or poor countries, especially those in island microstates and Sub-Saharan Africa, the election-supervising agency does not have a website with which to disseminate data. In these cases, the best source of constituency-level electoral data is archives maintained by scholars and election lovers. Four such archives are commonly referenced in our District-Level Election Data on the Internet (DLED) list, and so we describe them in detail here.

The most well known is the Lijphart Election Archive (LEA), which UCSD Emeritus Professor of Political Science Arend Lijphart began compiling in the early 1980’s and has informed several of the most influential publications on electoral systems and their effects (e.g. Lijphart 1984, 1990, 1994; Lijphart and Grofman 1984, 1986; Cox 1997). Lijphart and his colleagues have compiled the results of approximately 350 elections in 27 countries. Most of this data, however, is only available in hard copy from the UCSD library, so its utility to scholars in other parts of the country is unfortunately limited. In addition, the LEA website contain links to some on-line data sources.

For elections results in post-communist Europe, the most complete source is the Project on Political Transformation and the Electoral Process in Post-Communist Europe, hosted by the University of Essex. Besides the election results cited below, the University of Essex website has a searchable database on electoral laws in the region and changes in them over time. The data is available as an Excel file or an SPSS data file. The only drawback of the data is its format; instead of the rows containing the districts and separate columns the support for each party in that district, the votes for each party in a district is presented as a separate row. Hence the data must be reformatted before they can be used for statistical analysis of voting patterns at the district level.

Psephos: Adam Carr’s Electoral Archive is maintained by Australian journalist Adam Carr and contains election results for 163 countries as well as many sub-national entities. While not all the results are at the district level, in many cases the Psephos archive is the only online source for electoral data (especially for small countries). Another strength of the Psephos archive is that while many national-sources remove the links to previous election results as the new elections approach, the Psephos archive does delete old-election results. This archive, however, has two small drawbacks. First, small parties are often compiled into an “Other” category that for some studies, especially those interested in representation, political fragmentation, or the success of ethnic or gender parties, is a potential problem. Second, like the University of Essex data the results in the Psephos archive are presented with each candidate/party in a given district presented in a unique row. As a result, scholars must convert the data to a spreadsheet and then data manipulation, often a considerable amount, is required to make analysis of trends at the district level possible. Despite these drawbacks, the Psephos archive is an invaluable source of electoral data cross-sectionally and over time. For parsimony’s sake I have only listed below the elections for which it is the only source.

The fourth major resource available to scholars is Scott Morgenstern’s District-Level Electoral Data Website, compiled by Duke political science professor Scott Morgenstern for a series of articles with Richard Potthoff (2004, n.d.) on the components of electoral volatility. These data are easily downloaded as Excel files. The main disadvantage of this source is that the results are limited to the major parties (as noted in the “comments” section of our DLED list).

While not an on-line data source, one other important source of district-level election results merits mention: Daniele Caramani’s reference volume Elections in Western Europe since 1815: Electoral Results by Constituencies, (London: Macmillan Reference, 2000). This volume contains detailed histories of changes in the electoral laws and party systems of all West-European democracies. In addition, a CD-Rom containing detailed election results at the constituency level for all parties in a given election accompanies the volume. The CD-Rom does not have information about district magnitudes, so use of the Caramani data analysis of vote-seat relationships or the effect of district magnitude on party-fragmentation and the success of small parties requires additional research. Unfortunately (but understandably), the licensing agreement for the Data Archive CD-Rom prohibits us from posting any of this data on-line. But this volume is easy to use and meticulously complete in its description of election-law changes (though Morgenstern and Pothoff 2004 document some data-entry errors).